The pattern of switching of Schizosaccharomyces pombe (S. pombe)cells is nonrandom when assayed by single cell pedigrees. After two consecutive asymmetric cell divisions, one in four "granddaughter" cells undergoes a mating-type switch. Previously, we showed that this pattern is due to mat1 imprinting that marks only one sister chromatid in a strand-specific manner, and is related to a site-specific, double-stranded DNA break at mat1. We now show that this imprint is a strand-specific, alkali-labile DNA modification at mat1. The DNA break is an artifact, created from the imprint during DNA purification. We also proposed and tested the model that mat1 is preferentially replicated by a centromere-distal origin(s), so that the strand-specific imprint occurs only during lagging-strand synthesis. Altering the origin of replication, inverting mat1, or introducing an origin of replication affects the imprinting and switching efficiencies in predicted ways. Two-dimensional gel analysis confirmed that mat1 is preferentially replicated by a centromere-distal origin(s). Thus, the DNA replication machinery may confer different developmental potential to sister cells. Our recent work has discovered the biochemical functions of the swi1 and swi3 genes. We found that swi1p and swi3p perform imprinting by pausing and terminating DNA replication at mat1. Our work shows that: 1) the factors swi1p and swi3p act by pausing the replication fork at the imprinting site, and 2) swi1p and swi3p are involved in termination at the mat1-proximal polar-terminator of replication (RTS1). We performed a genetic screen to identify these termination factors and identified an allele that separated the pausing/imprinting and termination functions of swi1p. Our results suggest that swi1p and swi3p promote imprinting in novel ways, both by pausing replication at mat1 and by terminating replication at RTS1. We also showed that Swi1 and Swi3 proteins form a complex in vivo and that both proteins bind to the RTS1 and the mat1 replication pause sites on the chromosome. Future studies will be designed to define the mechanism of imprinting. We have already identified a large number of mat1 mutations that affect imprinting. Molecular analysis of these mat1 mutations should help us define the mechanism of imprinting.